Sora might soon become more accessible within ChatGPT, possibly leading to a surge in deepfakes from the video generator. OpenAI’s Sora video generator is set to be integrated into ChatGPT, according to The Information. Currently, Sora is available on its own website or as a standalone app, lacking the popularity of ChatGPT. This integration would let users access Sora’s video capabilities directly in ChatGPT, similar to last year’s addition of image generation features.
Integrating Sora into ChatGPT could increase user numbers but may also exacerbate the wave of deepfakes from OpenAI’s video generator. When Sora first launched, users created realistic deepfakes of historical figures and videos with copyrighted material. Making Sora more accessible may lead to more deepfakes and potential bypasses of OpenAI’s safeguards, such as watermark removal on AI-generated videos. Users have been finding loophole techniques in AI image and video generators, including tweaking prompts to fool the AI.
The potential integration of Sora comes as OpenAI faces growing competition. Anthropic’s Claude has gained popularity, while ChatGPT has seen an increase in uninstalls. Anthropic has received public support by refusing a Pentagon order for military use, in contrast to OpenAI’s acceptance, which has driven users away from ChatGPT. Integrating Sora may be an effort to win back users.
However, The Information notes that Sora could increase OpenAI’s costs, possibly resulting in additional pricing changes for ChatGPT, which introduced ads on its least expensive plans recently.